1: \documentclass[12pt,preprint]{aastex}
2:
3: \newcommand{\neii}{[\ion{Ne}{2}]}
4: \newcommand{\neiii}{[\ion{Ne}{3}]}
5: \newcommand{\nev}{[\ion{Ne}{5}]}
6: \newcommand{\siii}{[\ion{S}{3}]}
7: \newcommand{\oi}{[\ion{O}{1}]}
8: \newcommand{\arii}{[\ion{Ar}{2}]}
9: \newcommand{\hii}{H$_{2}$}
10: \newcommand{\feii}{[\ion{Fe}{2}]}
11:
12: \begin{document}
13:
14: \title{A Large Mass of H$_{2}$ in the Brightest Cluster Galaxy in
15: Zwicky~3146}
16:
17: \author{
18: E.~Egami\altaffilmark{1},
19: G.~H.~Rieke\altaffilmark{1},
20: D. Fadda\altaffilmark{2},
21: and
22: D.~C.~Hines\altaffilmark{3}
23: }
24:
25: \altaffiltext{1}{Steward Observatory, University of Arizona, 933
26: N. Cherry Avenue, Tucson, AZ~85721}
27:
28: \altaffiltext{2}{NASA Herschel Science Center, Caltech, MC 100-22, 770
29: S.\ Wilson Avenue, Pasadena, CA 91125}
30:
31: \altaffiltext{3}{Space Science Institute, 4750 Walnut Street, Suite
32: 205, Boulder, CO~80301}
33:
34: \begin{abstract}
35:
36: We present the {\em Spitzer}/IRS mid-infrared spectrum of the
37: infrared-luminous ($L_{IR}=4 \times 10^{11} L_{\sun}$) brightest
38: cluster galaxy (BCG) in the X-ray-luminous cluster Z3146 ($z=0.29$).
39: The spectrum shows strong aromatic emission features, indicating that
40: the dominant source of the infrared luminosity is star formation. The
41: most striking feature of the spectrum, however, is the exceptionally
42: strong molecular hydrogen (\hii) emission lines, which seem to be
43: shock-excited. The line luminosities and inferred warm \hii\ gas mass
44: ($\sim 10^{10} M_{\sun}$) are 6 times larger than those of NGC~6240,
45: the most \hii-luminous galaxy at $z \la 0.1$. Together with the large
46: amount of cold \hii\ detected previously ($\sim 10^{11} M_{\sun}$),
47: this indicates that the Z3146 BCG contains disproportionately large
48: amounts of both warm and cold \hii\ gas for its infrared luminosity,
49: which may be related to the intracluster gas cooling process in the
50: cluster core.
51:
52: \end{abstract}
53:
54: \keywords{galaxies: clusters: general --- cooling flows --- galaxies:
55: cD --- galaxies: active --- infrared: galaxies}
56:
57: \section{Introduction}
58:
59: Zwicky~3146 (Z3146)\footnote{The NASA/IPAC Extragalactic Database
60: (NED) name of this cluster is ZwCl1021.0+0426.} is an X-ray-luminous
61: cluster of galaxies at a redshift of 0.29. Its soft (0.1--2.4 keV)
62: X-ray luminosity \footnote{Throughout the paper, we adopt the
63: cosmological parameters of $\Omega_{M}=0.3$, $\Omega_{\Lambda}=0.7$,
64: and $H_{0}=70$ km s$^{-1}$.} of $2 \times 10^{45}$ erg s$^{-1}$
65: \citep{Ebeling98,Boehringer00} is one of the largest known, and based
66: on the {\em ROSAT} data, Z3146 was thought to exhibit one of the most
67: massive cluster cooling flows \citep{Edge94}. Our analysis of {\em
68: Chandra} data confirms that the gas cooling rate is indeed large with
69: an estimated mass deposition rate of $\sim 300$ $M_{\sun}$ yr$^{-1}$
70: within a 50 kpc radius from the cluster center \citep{Egami06}.
71:
72: The brightest cluster galaxy (BCG) is exceptionally active, showing a
73: blue continuum and luminous nebular emission lines \citep{Allen92,
74: Hicks05} and an infrared luminosity of $4 \times 10^{11}$ L$_{\sun}$
75: \citep{Egami06} . The mid-infrared spectral energy distribution (SED)
76: suggests that the source of the large infrared luminosity is star
77: formation at a rate of $\sim$70 $M_{\sun}$ yr$^{-1}$, broadly
78: consistent with the mass deposition rate mentioned above. This may
79: indicate that the active star formation is caused by cooling cluster
80: gas accreting onto the BCG, i.e., the cooling flow \citep{Cowie77,
81: Fabian77}. At the same time, the BCG seems to contain a radio AGN
82: that may contribute to its luminosity \citep{Egami06}.
83:
84: To investigate the mid-infrared spectral properties in detail, we have
85: observed the BCG in Z3146 using the Infrared Spectrograph (IRS; Houck
86: et al. 2004) on board the {\em Spitzer} Space Telescope
87: \citep{Werner04}.
88:
89: \section{Observations and Data Reduction}
90:
91: The IRS spectra of the BCG in Z3146 covering 5--38 $\mu$m were taken
92: in the staring mode with the two low-resolution modules, Short-Low
93: (SL; $\Delta\lambda = 5.2 - 14.5 \mu$m) and Long-Low (LL;
94: $\Delta\lambda = 14.0 - 38.0 \mu$m), with a resolving power of $R \sim
95: 64 - 128$. The SL spectra were taken for 8 cycles of 60 s ramp
96: durations for both the 2nd (SL2: 5.2--8.7 $\mu$m) and 1st (SL1:
97: 7.4--14.5 $\mu$m) orders while the LL spectra were taken for 8 cycles
98: of 120 s ramp durations for the 2nd order (LL2: 14.0--21.3$\mu$m) and
99: 6 cycles for the 1st order (LL1: 19.5-- 38.0 $\mu$m). The resultant
100: integration times are 960 s (SL1 and SL2), 1920 s (LL2), and 1440 s
101: (LL1), respectively, after coadding the spectra taken at the two nod
102: positions. The slit width is 3\farcs7 for the SL and 10\farcs7 for
103: the LL module.
104:
105: Starting with the reduced and coadded two dimensional spectral image
106: at each nod position, we performed sky subtraction in two passes,
107: first by taking the difference between the two nod positions (spectral
108: images targetting different orders were paired to avoid both the
109: positive and negative spectra falling in the same order) and second by
110: fitting and subtracting a linear fit to the background along each row
111: in the cross-dispersed direction (only fitting the pixels well away
112: from the source spectrum). Then, pixels with an abnormal signal near
113: the source spectrum were manually masked out from the subsequent
114: processing. From the background-subtracted and cleaned spectral
115: images, the source spectrum was extracted at each nod position using
116: the program
117: SPICE\footnote{\url{http://ssc.spitzer.caltech.edu/postbcd/spice.html}}
118: with extraction apertures reduced to 75\% of the default widths. All
119: the spectra were combined by resampling them with a uniform resolving
120: power of 130 over the whole wavelength range. Finally, the combined
121: spectrum was multiplied by 1.4 to match the IRAC 8.0 $\mu$m and MIPS
122: 24 $\mu$m photometry.
123:
124: \section{Results}
125:
126: The final combined IRS spectrum of the BCG in Z3146 is shown in
127: Figure~\ref{irs}. The measured line fluxes and rest-frame equivalent
128: widths are listed in Table~\ref{flux}. Along the spatial direction,
129: the lines and continuum are unresolved (FWHM$\sim$2 pixels), which
130: sets an upper limit of 3\farcs6 in SL and 10\farcs2 in LL,
131: respectively, on the size of the mid-infrared emitting region. This
132: is consistent with the fact that the Z3146 BCG is barely resolved at 8
133: $\mu$m with a spatial resolution of 2\arcsec\ \citep{Egami06}.
134:
135: Strong aromatic (``PAH'') emission features are detected at 6.2, 7.7,
136: 8.6, and 11.3 $\mu$m. These features indicate that the mid-infrared
137: (and therefore presumably the far-infrared) luminosity of this galaxy
138: has a significant component of active star formation.
139:
140: The spectrum also shows exceptionally strong emission lines from
141: molecular hydrogen (\hii). We have securely detected the 0--0 S(1),
142: S(2), S(3), S(4), and S(5) pure rotational lines. Although the 0--0
143: S(0) line is within the observed wavelength range ($\lambda_{obs} =
144: 36.4 \mu$m), the noise in the LL1 band increases dramatically beyond
145: 35 $\mu$m, preventing us from setting a meaningful upper limit on the
146: line flux. The 0--0 S(6) line is blended with the bright 6.2 $\mu$m
147: PAH feature nearby.
148:
149: A few atomic fine-structure emission lines are also detected in the
150: spectrum such as \neii\ 12.8 $\mu$m and \neiii\ 15.5 $\mu$m. The
151: \siii\ 18.7 $\mu$m line is possibly detected as well, but because of
152: the low signal-to-noise ratio, we treat it as a non-detection here.
153: Also, no high-ionization line (e.g., \nev\ 15.6$\mu$m) is seen.
154: The small \neiii/\neii\ line ratio of 0.2 is typical of a starburst
155: galaxy \citep{Thornley00} although we cannot rule out a significant
156: contribution to these neon lines from a shock-excited gas as
157: discussed below.
158:
159: There are possible detections of the \feii\ lines at 5.3 and 26.0
160: $\mu$m. However, a longer integration time is needed to confirm the
161: reality of these lines.
162:
163: \section{Discussion}
164:
165: \subsection{PAH Features}
166:
167: In terms of the strength of the 6.2 $\mu$m PAH feature, the Z3146 BCG
168: appears to be a typical starburst-dominated infrared luminous galaxy.
169: The rest-frame equivalent width of the feature (0.67 $\mu$m) as well
170: as its luminosity ratio to the total infrared luminosity ($5 \times
171: 10^{-3}$) are both within the range found for starburst galaxies,
172: $\sim 0.6\pm0.1$ $\mu$m for the former \citep{Weedman05} and $(6.3 \pm
173: 3.2) \times 10^{-3}$ for the latter \citep{Peeters04}, respectively.
174: Given the absence of any AGN-related features in the spectrum, the AGN
175: contribution to the infrared luminosity is likely to be small.
176:
177: \subsection{Molecular Hydrogen Lines}
178:
179: %\begin{figure*}
180: % \hspace*{2cm}\includegraphics[angle=90,scale=0.55]{fig2.eps}
181: % \caption{H2 excitation diagram. \label{h2}}
182: %\end{figure}
183:
184: The most striking feature of the spectrum is the exceptionally strong
185: emission lines from molecular hydrogen. The luminosities of the pure
186: rotational lines are on average 6.3 times larger than those of
187: NGC~6240, the most \hii-luminous galaxy at $z \la 0.1$.
188:
189: The \hii\ line strengths are even more remarkable when compared with
190: the infrared luminosity of this BCG. Figure~\ref{lum_ratio}a plots
191: the line luminosity of the \hii\ 0--0 S(1) emission line against the
192: infrared luminosity of various galaxies. The \hii\ 0--0 S(1)/infrared
193: luminosity ratio of this BCG is extreme (0.25\%), almost an order of
194: magnitude larger than the ratio for NGC~6240 (0.03\%), and a factor of
195: 50 above the trend line for the other galaxies (0.005\%).
196:
197: Figure~\ref{h2} shows the \hii\ level population in the Z3146 BCG.
198: For comparison, the level population in NGC~6240 is overplotted after
199: being scaled by a factor of 6.3. The level populations are quite
200: similar in shape between the two galaxies in the lowest energy states
201: ($v=0$; $J=3-7$). These lowest energy states are almost always
202: thermalized (i.e., collisionally excited), and therefore the
203: similarity suggests that although the total mass of warm \hii\ differs
204: between the two galaxies, the fractional \hii\ mass distribution as a
205: function of temperature is similar up to a gas temperature of
206: $\sim$960 K (i.e., the \hii\ excitation temperature derived for the
207: $v=0$, $J=6$ and 7 states using the 0--0 S(4) and S(5) lines).
208:
209:
210: Assuming that this similarity of the level population extends down to
211: the $v=0, J=2$ state, from which the 0--0 S(0) line originates, we
212: estimate the total warm \hii\ mass in the Z3146 BCG to be $\sim
213: 10^{10} M_{\sun}$, i.e., 6.3 times larger than that of NGC~6240 ($1.6
214: \times 10^{9} M_{\sun}$ by \citet{Armus06}). This mass is dominated
215: by the $\sim$160 K component, a temperature derived from the 0--0
216: S(0)/0--0 S(1) line ratio measured for NGC~6240. The cold \hii\ mass
217: derived from the CO observation is also large, $\sim 10^{11} M_{\sun}$
218: \citep{Edge03}, but the fraction of warm \hii\ mass is high,
219: $\sim$10\%, similar to the 15\% fraction found in NGC~6240
220: \citep{Armus06}.
221:
222: Figure~\ref{h2} also shows that the \hii\ level population of the
223: Z3146 BCG in the $v=1$ vibrational state is highly suppressed (by a
224: factor of 30) with respect to that in the $v=0$ state when the
225: NGC~6240 $v=0$ level population is used as a template. Although part
226: of this suppression is due to the slit loss with the near-infrared
227: spectroscopy as well as internal dust extinction, these effects, even
228: when combined, are unlikely to be sufficient. For example, the size
229: of the mid-infrared emitting region is 2\arcsec\ at most as already
230: mentioned, so the 1\farcs5 slit used to measure the $v=1-0$
231: near-infrared \hii\ lines \citep{Edge02} should capture most of the
232: total line fluxes. Also, in this BCG, the extinction derived from the
233: Balmer decrement is $A_{V}=0.6 $ mag \citep{Crawford99}, and when
234: corrected for this amount, the H$\alpha$-derived star formation rate
235: agrees well with the infrared-derived star formation rate
236: \citep{Egami06}, suggesting that the extinction cannot be much larger.
237: In fact, near-/mid-infrared \hii\ emission lines are known to show
238: little extinction in infrared-luminous galaxies
239: \citep{Goldader97,Higdon06}.
240:
241: This suppression of the $v=1$ state suggests that the $v=1$ level
242: population in the Z3146 BCG is subthermal as is the case with NGC~6240
243: \citep{Egami98,Lutz03}. Although the cause of such a subthermal level
244: population is not clear, one possibility is that the \hii\ molecules
245: are excited by a continuous (i.e., C-type) shock, which can produce
246: subthermal level populations in the low vibrational states
247: \citep[e.g.,][]{Kaufman96,LeBourlot02}. Note that alternative
248: excitation mechanisms such as a jump (J-type) shock or far-UV
249: radiation would produce a $v=1$ level population at or above the
250: thermal level with respect to the $v=0$ state
251: \citep[e.g.,][]{Burton92}. Figure~\ref{h2} shows that only a factor
252: of $\sim$7 scaling is required to bring up the $v=1$ level population
253: in the Z3146 BCG to that of NGC~6240. This correction factor is much
254: easier to account for than the factor of 30 mentioned above, which
255: would be required if the $v=1$ state is thermalized. Shock excitation
256: is also consistent with the previous near-infrared spectroscopic
257: studies of nearby cooling-flow cluster BCGs
258: \citep[e.g.,][]{Jaffe01,Wilman02}.
259:
260: \subsection{Atomic Fine-Structure Lines}
261:
262: The \neii\ 12.8 $\mu$m line of this BCG is also strong.
263: Figure~\ref{lum_ratio}b plots the \neii\ 12.8 $\mu$m luminosity
264: against the infrared luminosity for a sample of nearby infrared-bright
265: galaxies. As shown by \citet{Sturm02}, there is a good correlation
266: between the two luminosities among the starburst and Seyfert galaxies,
267: and the Z3146 BCG again stands out by having an exceptionally strong
268: \neii\ 12.8 $\mu$m line for its infrared luminosity (0.75\%),
269:
270: This overluminous \neii\ emission line may also arise from a shock.
271: However, the C-type shock invoked above to explain the \hii\ level
272: population cannot produce a substantial amount of the \neii\ line
273: because the shocked gas in a C-type shock is largely neutral and
274: molecular. To produce the \neii\ line, the shock must be a fast ($\ga
275: 50$ km s$^{-1}$) J-type shock, that can dissociate and ionize the
276: shocked gas \citep[e.g.,][]{Hollenbach89}. The Z3146 BCG has an
277: H$\alpha$ velocity width of $\sim600$ km s$^{-1}$ \citep{Allen92}, so
278: such a fast shock is plausible.
279:
280: This probably means that various types of shocks may coexist in the
281: Z3146 BCG. A similar conclusion was reached for NGC~6240 by
282: \citet{Lutz03} to account for the strong \oi\ 63 $\mu$m line, which is
283: also a sign of a J-type shock. Indeed, such a coexistence of the C-
284: and J-type shocks is also seen in some Galactic outflow sources, in
285: which both strong \hii\ and \neii\ lines are detected
286: \citep[e.g.,][]{Lefloch03}. Shock models can produce a wide range of
287: \neiii/\neii\ line ratios depending on various shock conditions, and
288: therefore it is possible to produce the starburst like \neiii/\neii\
289: ratio with a strong shock \citep[e.g.,][]{Binette85, Hartigan87}.
290: \feii\ lines may also be used to constrain the shock conditions if
291: their strengths are measured accurately.
292:
293: With respect to the \neii\ 12.8 $\mu$m line, the \siii\ 18.7 $\mu$m
294: line is abnormally weak (\siii/\neii\ $<$0.05). Such a low
295: \siii/\neii\ ratio is often seen with ultraluminous infrared galaxies
296: (ULIRGs), indicating that the gas phase abundance of sulfur is
297: significantly below solar \citep{Genzel98}. Starburst galaxies also
298: show this sulphur ``deficiency'' although the \siii/\neii\ ratios are
299: less extreme, and this may suggest that the sulphur is depleted onto
300: dust grains in high-metallicity star-forming systems \citep{Verma03}.
301:
302:
303:
304: \section{Conclusions}
305:
306: We present the {\em Spitzer}/IRS mid-infrared spectrum of the
307: infrared-luminous BCG in the X-ray luminous cluster Z3146. The strong
308: aromatic features indicate that the dominant source of the infrared
309: luminosity is star formation. The most striking feature of the
310: spectrum, however, is the exceptionally strong \hii\ emission lines.
311: To our knowledge, this BCG has the most luminous \hii\ pure-rotational
312: emission lines and most massive warm \hii\ gas component ($\sim
313: 10^{10} M_{\sun}$) known. Together with the large amount of cold
314: \hii\ detected previously ($\sim 10^{11} M_{\sun}$), this indicates
315: that the Z3146 BCG contains disproportionately large amounts of both
316: warm and cold \hii\ gas for its infrared luminosity. Since the parent
317: cluster Z3146 has an exceptionally large gas cooling rate in the core,
318: these massive warm and cold \hii\ components may be related to the
319: intracluster gas cooling process.
320:
321: \acknowledgments
322:
323: We would like to thank Dr.\ G.~Neugebauer and the anonymous referee
324: for their comments, which improved the paper. This work is based on
325: observations made with the {\em Spitzer} Space Telescope, which is
326: operated by the Jet Propulsion Laboratory, California Institute of
327: Technology under a contract with NASA (contract number \#1407).
328: Support for this work was provided by NASA through an award issued by
329: JPL/Caltech (contract number \#1255094).
330:
331: \begin{thebibliography}{}
332:
333: \bibitem[Allen et al.(1992)]{Allen92} Allen, S.~W., et al.\
334: 1992, \mnras, 259, 67
335:
336: \bibitem[Armus et al.(2006)]{Armus06} Armus, L., et al.\ 2006,
337: \apj, 640, 204
338:
339: \bibitem[Binette et al.(1985)]{Binette85} Binette, L., Dopita,
340: M.~A., \& Tuohy, I.~R.\ 1985, \apj, 297, 476
341:
342: \bibitem[B{\" o}hringer et al.(2000)]{Boehringer00} B{\" o}hringer,
343: H., et al.\ 2000, \apjs, 129, 435
344:
345: \bibitem[Burton(1992)]{Burton92} Burton, M.~G.\ 1992,
346: Australian Journal of Physics, 45, 463
347:
348: \bibitem[Cowie \& Binney(1977)]{Cowie77} Cowie, L.~L., \&
349: Binney, J.\ 1977, \apj, 215, 723
350:
351: \bibitem[Crawford et al.(1999)]{Crawford99} Crawford, C.~S., Allen,
352: S.~W., Ebeling, H., Edge, A.~C., \& Fabian, A.~C.\ 1999, \mnras, 306,
353: 857
354:
355: \bibitem[Ebeling et al.(1998)]{Ebeling98} Ebeling, H., Edge,
356: A.~C., Bohringer, H., Allen, S.~W., Crawford, C.~S., Fabian, A.~C., Voges,
357: W., \& Huchra, J.~P.\ 1998, \mnras, 301, 881
358:
359: \bibitem[Edge et al.(1994)]{Edge94} Edge, A.~C., Fabian,
360: A.~C., Allen, S.~W., Crawford, C.~S., White, D.~A., Bohringer, H., \&
361: Voges, W.\ 1994, \mnras, 270, L1
362:
363: \bibitem[Edge et al.(2002)]{Edge02} Edge, A.~C., Wilman, R.~J.,
364: Johnstone, R.~M., Crawford, C.~S., Fabian, A.~C., \& Allen, S.~W.\
365: 2002, \mnras, 337, 49
366:
367: \bibitem[Edge \& Frayer(2003)]{Edge03} Edge, A.~C., \& Frayer,
368: D.~T.\ 2003, \apjl, 594, L13
369:
370: \bibitem[Egami(1998)]{Egami98} Egami, E.\ 1998, IAU Symp.~184:
371: The Central Regions of the Galaxy and Galaxies, 184, 95
372:
373: \bibitem[Egami et al.(2006)]{Egami06} Egami, E., et al.\
374: 2006, \apj, 647, 922
375:
376: \bibitem[Fabian \& Nulsen(1977)]{Fabian77} Fabian, A.~C., \&
377: Nulsen, P.~E.~J.\ 1977, \mnras, 180, 479
378:
379: \bibitem[Genzel et al.(1998)]{Genzel98} Genzel, R., et al.\
380: 1998, \apj, 498, 579
381:
382: \bibitem[Goldader et al.(1997)]{Goldader97} Goldader, J.~D.,
383: Joseph, R.~D., Doyon, R., \& Sanders, D.~B.\ 1997, \apjs, 108, 449
384:
385: %\bibitem[Gordon et al.(2005)]{Gordon05} Gordon, K.~D., et al.\
386: %2005, \pasp, 117, 503
387:
388: \bibitem[Hartigan et al.(1987)]{Hartigan87} Hartigan, P., Raymond,
389: J., \& Hartmann, L.\ 1987, \apj, 316, 323
390:
391: \bibitem[Higdon et al.(2006)]{Higdon06} Higdon, S.~J.~U., Arms, L.,
392: Higdon, J.~L., Soifer, B.~T., \& Spoon, H.~W.~W. 2006, ApJ, in press
393: (astro-ph/0605359)
394:
395: \bibitem[Hollenbach \& McKee(1989)]{Hollenbach89} Hollenbach, D., \&
396: McKee, C.~F.\ 1989, \apj, 342, 306
397:
398: \bibitem[Jaffe et al.(2001)]{Jaffe01} Jaffe, W., Bremer,
399: M.~N., \& van der Werf, P.~P.\ 2001, \mnras, 324, 443
400:
401: \bibitem[Kaufman \& Neufeld(1996)]{Kaufman96} Kaufman, M.~J., \&
402: Neufeld, D.~A.\ 1996, \apj, 456, 611
403:
404: \bibitem[Le Bourlot et al.(2002)]{LeBourlot02} Le Bourlot, J.,
405: Pineau des For{\^e}ts, G., Flower, D.~R., \& Cabrit, S.\ 2002,
406: \mnras, 332, 985
407:
408: \bibitem[Lefloch et al.(2003)]{Lefloch03} Lefloch, B.,
409: Cernicharo, J., Cabrit, S., Noriega-Crespo, A., Moro-Mart{\'{\i}}n,
410: A., \& Cesarsky, D.\ 2003, \apjl, 590, L41
411:
412: \bibitem[Lutz et al.(2003)]{Lutz03} Lutz, D., Sturm, E.,
413: Genzel, R., Spoon, H.~W.~W., Moorwood, A.~F.~M., Netzer, H., \& Sternberg,
414: A.\ 2003, \aap, 409, 867
415:
416: \bibitem[Hicks \& Mushotzky(2005)]{Hicks05} Hicks, A.~K., \&
417: Mushotzky, R.\ 2005, \apjl, 635, L9
418:
419: \bibitem[Peeters et al.(2004)]{Peeters04} Peeters, E., Spoon,
420: H.~W.~W., \& Tielens, A.~G.~G.~M.\ 2004, \apj, 613, 986
421:
422: \bibitem[Rieke et al.(2004)]{Rieke04} Rieke, G.~H., et al.\
423: 2004, \apjs, 154, 25
424:
425: \bibitem[Rigopoulou et al.(2002)]{Rigopoulou02} Rigopoulou, D.,
426: Kunze, D., Lutz, D., Genzel, R., \& Moorwood, A.~F.~M.\ 2002, \aap, 389,
427: 374
428:
429: \bibitem[Sanders et al.(2003)]{Sanders03} Sanders, D.~B.,
430: Mazzarella, J.~M., Kim, D.-C., Surace, J.~A., \& Soifer, B.~T.\ 2003,
431: \aj, 126, 1607
432:
433: \bibitem[Sturm et al.(2002)]{Sturm02} Sturm, E., Lutz, D.,
434: Verma, A., Netzer, H., Sternberg, A., Moorwood, A.~F.~M., Oliva, E., \&
435: Genzel, R.\ 2002, \aap, 393, 821
436:
437: \bibitem[Sugai et al.(1997)]{Sugai97} Sugai, H., Malkan, M.~A., Ward,
438: M.~J., Davies, R.~I., \& McLean, I.~S.\ 1997, \apj, 481, 186
439:
440: \bibitem[Thornley et al.(2000)]{Thornley00} Thornley, M.~D.,
441: Schreiber, N.~M.~F., Lutz, D., Genzel, R., Spoon, H.~W.~W., Kunze, D., \&
442: Sternberg, A.\ 2000, \apj, 539, 641
443:
444: \bibitem[Verma et al.(2003)]{Verma03} Verma, A., Lutz, D., Sturm, E.,
445: Sternberg, A., Genzel, R., \& Vacca, W.\ 2003, \aap, 403, 829
446:
447: \bibitem[van der Werf(1996)]{vdWerf96} van der Werf, P.~P.\ 1996, ASSL
448: Vol.~206: Cold Gas at High Redshift, 37
449:
450: \bibitem[Weedman et al.(2005)]{Weedman05} Weedman, D.~W., et al.\
451: 2005, \apj, 633, 706
452:
453: \bibitem[Werner et al.(2004)]{Werner04} Werner, M.~W., et al.\
454: 2004, \apjs, 154, 1
455:
456: \bibitem[Wilman et al.(2002)]{Wilman02} Wilman, R.~J., Edge,
457: A.~C., Johnstone, R.~M., Fabian, A.~C., Allen, S.~W., \& Crawford, C.~S.\
458: 2002, \mnras, 337, 63
459:
460: \end{thebibliography}
461:
462: \clearpage
463:
464: \begin{deluxetable}{lccc}
465: \tablecaption{Z3146 BCG line fluxes \label{flux}}
466: \tablewidth{0pt}
467: %\tablewidth{2.5in}
468: \tablehead{
469: \colhead{Line/Feature} & \colhead{$\lambda_{\rm rest}$} & \colhead{Flux} &
470: \colhead{EW$_{rest}$} \\
471: \colhead{} & \colhead{($\mu$m)} & \colhead{(10$^{-17}$ W m$^{-2}$)} &
472: \colhead{($\mu$m)}
473: }
474: \startdata % EWobs
475: PAH & 6.2 & 3.0$\pm$0.2 & 0.67$\pm$0.05 \\ %0.87+-0.06
476: %H$_{2}$ \phn $0-0$ S(5)& 6.9 & 2.7$\pm$0.1 & 0.64$\pm$0.03 \\ %0.83+-0.04
477: H$_{2}$ \phn $0-0$ S(5) & 6.9 & 2.1$\pm$0.1 & 0.49$\pm$0.02 \\ %
478: \arii\tablenotemark{a} & 6.99 & $\sim$0.6 & $\sim$0.15 \\ %
479: PAH & 7.7 & 6.9$\pm$0.2 & 1.72$\pm$0.06 \\ %2.22+-0.08
480: H$_{2}$ \phn $0-0$ S(4) & 8.0 & 0.8$\pm$0.1 & 0.19$\pm$0.02 \\ %0.25+-0.03
481: PAH & 8.6 & 1.2$\pm$0.2 & 0.31$\pm$0.04 \\ %0.40+-0.05
482: H$_{2}$ \phn $0-0$ S(3) & 9.7 & 2.3$\pm$0.1 & 1.72$\pm$0.05 \\ %2.23+-0.06
483: PAH & 11.3 & 1.3$\pm$0.1 & 0.47$\pm$0.07 \\ %0.60+-0.09
484: H$_{2}$ \phn $0-0$ S(2) & 12.3 & 1.2$\pm$0.1 & 0.26$\pm$0.02 \\ %0.34+-0.02
485: \neii & 12.8 & 4.2$\pm$0.1 & 1.34$\pm$0.03 \\ %1.73+-0.04
486: \neiii & 15.6 & 0.8$\pm$0.1 & 0.30$\pm$0.04 \\ %0.39+-0.05
487: H$_{2}$ \phn $0-0$ S(1) & 17.0 & 1.4$\pm$0.1 & 0.55$\pm$0.04 \\ %0.71+-0.05
488: \siii & 18.7 & $<0.4$ & $< 0.2$ \\ %0.10+-0.07
489: \enddata
490:
491: \tablenotetext{a}{The \arii\ 6.99 $\mu$m line was not detected
492: directly, but its presence was set to the maximum level consistent
493: with the S(5) line shape.}
494:
495: \end{deluxetable}
496:
497: \clearpage
498:
499: \begin{figure}
500: \includegraphics[angle=90,scale=0.65]{f1.eps}
501: \caption{Combined IRS spectrum of the BCG in Z3146. The two points
502: with error bars show the IRAC 8.0$\mu$m and MIPS 24 $\mu$m photometry
503: from \citet{Egami06}. The IRS spectrum has been scaled ($\times 1.4$)
504: to match these photometry points when the spectrum is integrated over
505: the corresponding passbands. \label{irs}}
506: \end{figure}
507:
508: \clearpage
509:
510: \begin{figure}
511: \includegraphics[angle=90,scale=0.65]{f2.eps}
512: \caption{(a) \hii\ 0--0 S(1) 17 $\mu$m line luminosity vs. infrared
513: luminosity ($L_{IR}$). The solid and dashed lines correspond to
514: $L_{0-0S(1)} = $ 0.005 and 0.25\% of $L_{IR}$. The \hii\ line
515: luminosities were taken from \citet{Armus06} (NGC~6240),
516: \citet{Rigopoulou02}, and \citet{Higdon06}. Note that the \hii\
517: luminosity correlates with the infrared luminosity, so the sample
518: studied by \citet{Higdon06}, which extends beyond a redshift of 0.1
519: and therefore includes more infrared-luminous galaxies, contains
520: galaxies more \hii\ luminous than NGC~6240; (b) \neii\ 12.8 $\mu$m
521: line luminosity vs.\ $L_{IR}$. The solid and dashed lines correspond
522: to $L_{[Ne II]} = $ 0.06 and 0.75\% of $L_{IR}$. The \neii\ line
523: luminosities were taken from \citet{Genzel98} while the infrared
524: luminosities were taken from \citet{Sanders03}. \label{lum_ratio}}
525: \end{figure}
526:
527: \clearpage
528:
529: \begin{figure}
530: \includegraphics[angle=90,scale=0.65]{f3.eps}
531: \caption{\hii\ excitation diagram showing the level population of warm
532: \hii\ in the Z3146 BCG. The level population in the ground
533: vibrational state ($v=0$) is based on our observation while the $v=1$
534: level population was derived from the data by \citet{Edge02}. The X
535: axis is the \hii\ energy level expressed in temperature ($E_{i}$ is
536: the energy level of the $i$-th state; $k$ is the Boltzmann constant).
537: In the Y axis, $N_{i}$ refers to the {\em total} number of \hii\ in
538: the $i$-th energy state while $g_{i}$ is the statistical weight of
539: that state (ortho-para ratio was set to 3:1). Assuming that the \hii\
540: lines are optically thin, $N_{i}$ was calculated as $N_{i} =
541: L_{i}/(A_{i} h \nu_{i})$, where $L_{i}$ is the observed line
542: luminosity, $A_{i}$ is the Einstein coefficient for that transition,
543: $h$ is the Planck constant, and $\nu_{i}$ is the frequency of the
544: line. In comparison, the \hii\ level population of NGC~6240 is also
545: shown based on the published data
546: \citep{vdWerf96,Sugai97,Lutz03,Armus06}. The data from
547: \citet{Sugai97} were plotted by assuming that the 1--0 S(1) line flux
548: agrees with that of \citet{vdWerf96}. No other scaling was applied.
549: The solid line connects the $v=0$ and 1 states in the Z3146 BCG while
550: the dotted line connects the $v=0$ states in NGC~6240. \label{h2}}
551: \end{figure}
552:
553: \end{document}
554:
555:
556: